The exponential expansion of Big Data in 7V's (velocity, variety, veracity, value, variability, and visualization) brings forth new challenges to security, reliability, availability, and privacy of these data sets. Traditional security techniques and algorithms fail to complement this big gigantic data. This paper aims to improve the recently proposed Atrain Distributed System (ADS) by incorporating new features, which will cater to the end-to-end availability and security aspects of the big data in the distributed system. The paper also integrates the concept of Software Defined Networking (SDN) in ADS to effectively control and manage the routing of the data item in the ADS. The storage of data items in the ADS is done based on the type of data (structured or unstructured), the capacity of the distributed system (or coach) and the distance of coach from the pilot computer (PC). In order to maintain the consistency of data and to eradicate the possible loss of data, the concept of "forward positive" and "backward positive" acknowledgment is proposed. Furthermore, we have incorporated "Twofish" cryptographic technique to encrypt the big data in the ADS. Issues like "data ownership," "data security, "data privacy" and data reliability" are pivotal while handling the big data. The current paper presents a framework for a privacy-preserved architecture for effectively handling the big data.